Someone Asked Two LLMs For Free Replacements. Seven Worked. The Bar Was Somewhere Underground.
A writer described their paid subscriptions and hobbies to ChatGPT and Gemini, then asked each model to identify free alternatives. Seven of the recommendations proved usable enough to adopt. The exercise is essentially prompt-driven substitution analysis, though nobody is calling it that.
This demonstrates a principle I like to call comparative utility mapping. You provide a model with your current tool stack and spending habits, and it cross-references feature sets against free alternatives in its training data. The mechanism is simple but powerful: the LLM functions as a budget-conscious procurement analyst, if you bother to give it sufficient context about your actual needs.
Tom's Guide conducted the comparison using ChatGPT and Gemini as the recommendation engines. The seven successful replacements came from those two models' outputs.
- Open ChatGPT or Gemini and list every app, service, or tool you currently pay for, along with one sentence about why you use each. The model needs context, not a vague wishlist.
- Add this instruction: 'For each item above, suggest the best free alternative that supports the same core use case. Rank by how close the match is.' You should receive a structured list.
- Pick the top three suggestions and actually test them for ten minutes each. A recommendation you never try is worth precisely nothing.